AIIMS, Google develop AI-powered app to expand skin care access: How your photo can help when there’s no derma nearby

The proposed app, Indus Derma, will also suggest next steps, including whether the condition requires urgent referral to a dermatologist, can be managed locally or needs further investigations. It will also provide general treatment suggestions and lifestyle advice, such as skincare measures and preventive guidance.

Dermatology-related problems are among the most frequently searched health issues online worldwide. “However, access to trained dermatologists remains limited, especially outside urban centres. Most dermatologists in India are concentrated in cities and many focus primarily on cosmetic and aesthetic practice, leaving semi-urban and rural populations underserved. As a result, patients in smaller towns and rural areas are often seen by general physicians or even non-physicians. AI can play a crucial role in empowering these healthcare providers by helping them triage cases effectively,” says Dr Somesh Gupta, professor of dermatology at AIIMS.

How the app works

Beyond diagnosis, the app will also suggest next steps, including whether the condition requires urgent referral to a dermatologist, can be managed locally or needs further investigations. Importantly, the app will not generate a legally valid prescription. Any medication suggestions provided by the AI will need to be verified and approved by a qualified medical professional. “This is a decision support tool, not a replacement for doctors,” Dr Gupta says.

Why AI is effective for dermatology

Dermatology is particularly well-suited for AI applications because it is largely a visual, pattern recognition specialty. “Most skin conditions are diagnosed based on appearance. This is where AI models, especially computer vision systems, perform extremely well,” Dr Gupta explains.

Even general-purpose AI models, such as ChatGPT, Gemini, and DeepSeek already perform at levels comparable to dermatologists for common skin conditions, he says. Specialised medical models like Google’s MedGemma sometimes perform even better in controlled settings.

Addressing an India-specific gap

One of the key challenges with existing global AI dermatology models is that they are trained predominantly on images of white skin. This data imbalance affects diagnostic accuracy for Indian patients.

To address this, AIIMS and Google are jointly developing Indian foundation models that will retrain global AI systems using Indian skin tones and dermatological conditions. “Skin colour significantly influences visual diagnosis. That’s why local data is essential,” says Dr Gupta.

How accurate is the app

The project is still undergoing development and validation. Regulatory approvals, accuracy benchmarks and extensive testing are required before public release. Dr Gupta says the team is aiming for an accuracy rate of 80 to 90 per cent, significantly higher than the approximately 40 per cent accuracy seen among non-specialists managing skin conditions.

Who will benefit the most

While experienced dermatologists may not need AI assistance, the app is expected to be especially useful for general physicians, healthcare workers and patients in areas with limited access to specialists. The app will help users understand the seriousness of their condition, decide whether they need to see a specialist urgently, and learn about basic care measures.

For patients, the app will also generate easy to understand information explaining what a condition is, why it occurs, and what general steps can help manage it. For doctors, the AI’s suggestions can be reviewed and incorporated into clinical decision making where appropriate.

While there is hope that a significant milestone could be reached by early next year, Dr Gupta says that a realistic launch timeline may be four to five months later than initial targets.

“The AIIMS Google dermatology app could mark a major shift in how skin conditions are assessed and managed in India, bringing expert level guidance closer to millions who currently lack access to specialist care,” he adds.